Imaging of Cerebral Blood Flow (CBF) is a powerful technique for the diagnostic evaluation of patients with dementia, stroke and epilepsy. A technique known as Arterial Spin Labeling MRI has demonstrated the ability to provide high quality images of CBF without radioactivity or injection. The advancement of this technique has been limited, however, by the destructive effects of patient motion, the poor image quality of the fast, echoplanar imaging technique used to combat these motion effects, and the inability to acquire images from the entire brain in a reasonable exam period. The applicants propose to overcome these limitations by refining a novel background suppression approach. Multiple inversion pulses will be used to reduce the intensity of uninteresting signals prior to image acquisition. This approach can reduce motion-related errors by a factor of 100 while preserving the CBF signal. Because of the dramatic reduction in motion-related errors, a superior, 3D fast spin echo imaging approach can be employed to provide CBF images from the whole brain in under 6 minutes with twice the sensitivity of earlier approaches. We propose to realize this potential of this approach by: 1. Optimizing the design of the RF pulses used for background suppression to minimize CBF signal loss and systematic errors near the top and bottom of the brain. 2. Using a nonlinear minimization strategy to optimize the timing of the inversions so as to achieve ideal background suppression 3. Developing a strategy for T1 quantification of brain tissue that will be compatible with the 3D fast spin echo sequence and which will be insensitive to the presence of cerebrospinal fluid(CSF). TI measurement is required for CBF measurement but can be contaminated by small amounts of CSF in the voxel. 4. Measuring the efficiency of the background suppressed sequences. Efficiency will be measured as a function of labeling plane location to guide the choice of parameters for subsequent applications. 5. Measuring the test-retest reliability of the optimized CBF imaging method in normal controls and patients with dementia and comparing it to unsuppressed methods. This information is needed for the design and interpretation of diagnostic tests, pharmaceutical evaluations and other studies employing CBF MRI. These developments will make reliable CBF imaging by Arterial Spin Labeling a widely applicable technique for diagnostic imaging.